Qualitative synthesis
Overview
Qualitative studies produce findings in words, and pooling them statistically would destroy the one thing they carry: meaning in context. A qualitative synthesis instead works from the findings upward — coding them, grouping them, and building descriptive and analytical themes or translations that answer a review question without flattening what the original authors saw. The choice of method decides what the output looks like: aggregated findings, developed themes, translated concepts, or an account of disagreement.
This skill covers that work and its choices: which synthesis approach fits which question, how coding proceeds honestly, how disagreement between studies is handled rather than averaged away, how study quality is appraised with a named tool, and how confidence in a finding is rated for evidence that has no confidence interval. The conduct of the search and screening is shared with systematic review conduct; this is what happens after the studies qualify.
Primary sources: https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/1471-2288-8-45 and https://casp-uk.net/casp-tools-checklists/.
When to use
- When synthesising qualitative or mixed-method studies' qualitative arms.
- When choosing between aggregation, thematic development, or translation.
- When disagreement between primary studies must be synthesised, not discarded.
- When rating confidence in qualitative findings.
When NOT to use
- For the review's search, screening, and protocol machinery (use ac-systematic-review-conduct).
- For statistical pooling of quantitative estimates (use ac-meta-analysis-and-synthesis).
- For writing the synthesis as a document (use ac-academic-writing).
Subskill Index
| Leaf |
Domain segment |
Scope (one line) |
subskills/approach-selection.md |
research-operations/approach-selection |
Matching the synthesis method to the question and its intent. |
subskills/coding-findings.md |
research-operations/coding-findings |
Line-by-line coding and the descent to concepts. |
subskills/meta-aggregation.md |
research-operations/meta-aggregation |
Assembling findings, categories, and syntheses without over-reach. |
subskills/thematic-synthesis.md |
research-operations/thematic-synthesis |
Descriptive themes, then analytical; the two-stage discipline. |
subskills/meta-ethnography.md |
research-operations/meta-ethnography |
First-, second-, third-order constructs and translation. |
subskills/refutational-synthesis.md |
research-operations/refutational-synthesis |
Disagreement as data: accounts that explain rather than average. |
subskills/appraising-quality.md |
research-operations/appraising-quality |
Named checklists and what appraisal does and does not license. |
subskills/confidence-in-findings.md |
research-operations/confidence-in-findings |
Rating qualitative evidence: CERQual and its domains. |
subskills/qualitative-record.md |
research-operations/qualitative-record |
The written record: approach, codes, decisions, and audit trail. |
Status and honest limits
Qualitative synthesis inherits its limits from the primary studies: thick description cannot be created where the studies are thin, and a theme that appears in twenty shallow studies is not deeper than one. Appraisal tools judge conduct, not importance; confidence ratings judge the finding's support, not whether the experience matters. Nothing here certifies what participants meant — it assembles what studies reported they meant, with each assembly step recorded so a reader can disagree with the assembly rather than discover it.
1---2name: ac-qualitative-synthesis3description: "Bringing qualitative studies together without destroying what makes them qualitative: choosing the approach, meta-aggregation and its findings-categories-syntheses levels, thematic synthesis and its descriptive then analytical themes, meta-ethnography and translations, refutational synthesis that treats disagreement as data, critical appraisal with a named checklist, CERQual confidence rating, and the written qualitative record. It explains how findings are produced from findings; it never reduces themes to counts. Use when asking \"qualitative synthesis\", \"thematic synthesis\", \"meta-ethnography\", \"meta-aggregation\", \"line by line coding\", \"third order constructs\", \"ENTREQ\", \"CASP qualitative checklist\", \"satur\", \"refutational synthesis\", \"qualitative evidence\". Not for the review's conduct machinery (ac-systematic-review-conduct), not for statistical pooling of quantitative studies (ac-meta-analysis-and-synthesis), not for writing the report (ac-academic-writing)."4license: MIT5metadata:6 source: https://skillmd.com/skills/jiayaoqijia/ac-qualitative-synthesis7 domain: research-operations8status: seedling9---1011# Qualitative synthesis1213## Overview1415Qualitative studies produce findings in words, and pooling them statistically would destroy the one thing they carry: meaning in context. A qualitative synthesis instead works from the findings upward — coding them, grouping them, and building descriptive and analytical themes or translations that answer a review question without flattening what the original authors saw. The choice of method decides what the output looks like: aggregated findings, developed themes, translated concepts, or an account of disagreement.1617This skill covers that work and its choices: which synthesis approach fits which question, how coding proceeds honestly, how disagreement between studies is handled rather than averaged away, how study quality is appraised with a named tool, and how confidence in a finding is rated for evidence that has no confidence interval. The conduct of the search and screening is shared with systematic review conduct; this is what happens after the studies qualify.1819Primary sources: <https://bmcmedresmethodol.biomedcentral.com/articles/10.1186/1471-2288-8-45> and <https://casp-uk.net/casp-tools-checklists/>.2021## When to use2223- When synthesising qualitative or mixed-method studies' qualitative arms.24- When choosing between aggregation, thematic development, or translation.25- When disagreement between primary studies must be synthesised, not discarded.26- When rating confidence in qualitative findings.2728## When NOT to use2930- For the review's search, screening, and protocol machinery (use ac-systematic-review-conduct).31- For statistical pooling of quantitative estimates (use ac-meta-analysis-and-synthesis).32- For writing the synthesis as a document (use ac-academic-writing).3334## Subskill Index3536| Leaf | Domain segment | Scope (one line) |37|---|---|---|38| `subskills/approach-selection.md` | `research-operations/approach-selection` | Matching the synthesis method to the question and its intent. |39| `subskills/coding-findings.md` | `research-operations/coding-findings` | Line-by-line coding and the descent to concepts. |40| `subskills/meta-aggregation.md` | `research-operations/meta-aggregation` | Assembling findings, categories, and syntheses without over-reach. |41| `subskills/thematic-synthesis.md` | `research-operations/thematic-synthesis` | Descriptive themes, then analytical; the two-stage discipline. |42| `subskills/meta-ethnography.md` | `research-operations/meta-ethnography` | First-, second-, third-order constructs and translation. |43| `subskills/refutational-synthesis.md` | `research-operations/refutational-synthesis` | Disagreement as data: accounts that explain rather than average. |44| `subskills/appraising-quality.md` | `research-operations/appraising-quality` | Named checklists and what appraisal does and does not license. |45| `subskills/confidence-in-findings.md` | `research-operations/confidence-in-findings` | Rating qualitative evidence: CERQual and its domains. |46| `subskills/qualitative-record.md` | `research-operations/qualitative-record` | The written record: approach, codes, decisions, and audit trail. |4748## Status and honest limits4950Qualitative synthesis inherits its limits from the primary studies: thick description cannot be created where the studies are thin, and a theme that appears in twenty shallow studies is not deeper than one. Appraisal tools judge conduct, not importance; confidence ratings judge the finding's support, not whether the experience matters. Nothing here certifies what participants meant — it assembles what studies reported they meant, with each assembly step recorded so a reader can disagree with the assembly rather than discover it.